From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 27697 invoked by alias); 17 Jan 2008 10:17:42 -0000 Received: (qmail 27688 invoked by uid 22791); 17 Jan 2008 10:17:42 -0000 X-Spam-Check-By: sourceware.org Received: from sibelius.xs4all.nl (HELO brahms.sibelius.xs4all.nl) (82.92.89.47) by sourceware.org (qpsmtpd/0.31) with ESMTP; Thu, 17 Jan 2008 10:17:25 +0000 Received: from brahms.sibelius.xs4all.nl (kettenis@localhost.sibelius.xs4all.nl [127.0.0.1]) by brahms.sibelius.xs4all.nl (8.14.1/8.14.0) with ESMTP id m0HAHLST015447; Thu, 17 Jan 2008 11:17:21 +0100 (CET) Received: (from kettenis@localhost) by brahms.sibelius.xs4all.nl (8.14.1/8.14.1/Submit) id m0HAHKoX024226; Thu, 17 Jan 2008 11:17:20 +0100 (CET) Date: Thu, 17 Jan 2008 10:17:00 -0000 Message-Id: <200801171017.m0HAHKoX024226@brahms.sibelius.xs4all.nl> From: Mark Kettenis To: guillaume.menant@geensys.com CC: gdb@sourceware.org In-reply-to: <14914940.post@talk.nabble.com> (message from Guillaume MENANT on Thu, 17 Jan 2008 01:18:51 -0800 (PST)) Subject: Re: 'g/G' GDB commands References: <473C1753.1090409@geensys.com> <20071115133842.GB19518@caradoc.them.org> <14437056.post@talk.nabble.com> <14914940.post@talk.nabble.com> Mailing-List: contact gdb-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-owner@sourceware.org X-SW-Source: 2008-01/txt/msg00143.txt.bz2 > Date: Thu, 17 Jan 2008 01:18:51 -0800 (PST) > From: Guillaume MENANT > > I've implemented the reply to the 'g' command but GDB sends me a nack ('-') > each time (3 times and then freeze). > > I'm working on a Atmel AT697E (Sparc V8) in which all registers printed with > "maint print registers" command exists except csr and dx registers (last > registers listed below). So I tried to send only supported registers (GDB > sends nack) and also tried to add '0' padding in order to fill the > non-implemented registers (same response). I'm not an expert on GDB's remote protocol, but here is some SPARC wisdom that might help you. The %dN registers are pseudo-registers (combining two single-precision %fN registers into a double-precision value), and should *not* be sent accross the line. The %csr register is indeed an implementation defined register (but a real one). We could define a special LEON2 subtarget in GDB, which doesn't have this register, but I think it would be easier if you just passed '0' for this one, since that would avoid all sorts of nasty configuration issues. Cheers, Mark